Generali France resolves 1.3 million calls autonomously and processes 2.1 million operations with Microsoft Copilot and agentic AI

Generali France, a major insurer covering savings, health, pension, personal, and property lines in France, faced a dual challenge: delivering faster, more personalized customer service while freeing employees from high-volume, repetitive work. In Property & Casualty specifically, claims spikes during climate events (droughts, storms) overwhelmed contact centers, and compliance obligations around beneficiary searches created persistent backlogs—15,705 unclaimed contracts in 2016 alone. Legacy RPA bots automated roughly 11% of processes but hit capability ceilings with unstructured data and nuanced document handling. Without augmentation, the status quo meant slow claims resolution, regulatory exposure, and deteriorating employee experience during peak demand periods.

Under its 'Boost 2027' strategic plan, Generali France deployed Microsoft 365 Copilot to 3,700 employees and extended Copilot Chat access to 100% of its workforce. Working with implementation partner Avanade, the company used Microsoft Copilot Studio and Azure OpenAI to build approximately 50 specialized AI agents integrated directly into existing workflows and information systems. Agents were deployed across helpdesk operations (drafting requests, tracking cases, escalating to specialists), a 24/7 voice assistant for claimants, home insurance subscription automation (replacing ~30 manual questions), and a 360° client-network visualization tool for compliance with beneficiary search obligations. An internal Cognitive Factory unit governs the automation program, with 17 use cases in production and roughly 30 more planned by end of 2025.

The deployment produced measurable outcomes across customer service, operations, and compliance:

• 1.3 million calls (30% of requests) resolved by customers without any human intervention via the voice assistant and self-service agents
• 2.1 million operations processed by RPA robots augmented with generative AI in 2024, including automatic recognition of 15+ document types
• 10% of crack-analysis claims during drought episodes now require no human review
• Unclaimed contracts fell from 15,705 (2016) to 802 (2023)
• Deaths identified and processed within a year rose from 33% (2017) to 88% (2023)
• 70% of employees actively use Copilot tools, averaging 15 prompts per user per week

Generali now leads Net Promoter Score rankings among French insurers.

• Generative AI is most valuable as a layer on top of existing RPA—rule-based bots hit ceilings with unstructured data, and LLM augmentation unlocks document classification, voice handling, and nuanced case routing that bots alone cannot perform.
• Universal access matters: giving 100% of staff access to AI tools (not just power users) combined with ambassador programs and structured training drove 70% active adoption.
• Compliance use cases in P&C have compounding returns—automating beneficiary search workflows reduced unclaimed contract backlogs by 95% over seven years.
• Climate-related claims spikes are a predictable pressure point; agentic voice assistants can absorb volume surges without proportional headcount increases.
• A dedicated internal center of excellence (Generali's Cognitive Factory) is essential for scaling from pilot to 50+ production agents without losing governance.